A system and method for configuring the display of a virtual workspace on multiple monitors connected to a single computing device based on images/frames captured by multiple cameras is described. A monitor orientation controller analyzes the frames to detect motion/movement within the captured scene and a corresponding centroid of the detected motion. The monitor orientation controller determines the positions of the monitors relative to each other based on the calculated centroids. Based on the relative determined positions of the monitors, the monitor orientation controller adjusts how the virtual workspace is displayed on the monitors. Other embodiments are also described.